A bit of a sad week for Stattraction, as StRAinE broke down and will require a rewrite of the code base in order to get going again. I’ve added a model status to the site so you can stay up to date if that’s your jam. However, it was a great week for STaRS, which pulled a perfect round of tips alongside great MAE and Bits metrics.
Brisbane are the team of the week this week after demolishing North, while the team that most improved their standing is West Coast. They pulled an average C after 3 straight failing grades to start the season.
It came to my attention this week that the change in rating automation broke again and was giving out random numbers. I’m not sure why this is happening after I fixed it last time, but alas. For the moment, I’m changing it to change in placing. Brisbane jump into first despite being 1-3 thanks to their phenomenal midfield rating. GWS are also killing it and retain their triple threat status.
GWS became outright premiership favourites this week, sneaking just in front of Melbourne. Brisbane jumped into the projected top 8 thanks to their big win over North, and the Bulldogs are perhaps-temporarily left out.
I think I’m going to get an email from Max “Mr. Squiggle” Barry over this one. Melbourne with a 54% chance of winning, by an average of -3 points? What is STaRS thinking?
Just to be clear, that’s not a freak error. I think tipping by negative margins is very funny and so I coded my model distinctly so it was a possibility. However, I didn’t think the predicted margin would ever get below -1. Further further inquiries can be directed here.
Otherwise, no real shockers this week. Most of the Squiggle models seem fairly united, too.
Stattraction 